The rebound in pFS firing rates on MD2 means that to restore RSU firing rates, homeostatic mechanisms must adjust excitation enough to precisely compensate both for the induction of LTD and for the rebound in pFS firing rates (which should recruit more inhibition onto RSUs). Because other (non-FS) classes of GABAergic interneurons cannot be cleanly

differentiated selleck chemicals llc from pyramidal neurons in these extracellular recordings, it is not clear whether all GABAergic neuron types express firing rate homeostasis, or if this is a property confined to pyramidal and FS cells. One puzzling question raised by the firing rate homeostasis hypothesis is how a homeostatic activity target can be implemented in a network that operates under very different sensory and modulatory conditions during different behavioral

states (Steriade and Timofeev, 2003 and Vyazovskiy et al., 2009). Because rodents sleep in short bouts interspersed with periods of active wake, our data provide a well-controlled opportunity to explore this question. One possibility is that neocortical networks have different set points during fundamentally different behavioral states. Ibrutinib cost Another possibility is that homeostatic regulation only constrains the activity of neurons in certain states (wake, for example), while firing rates during other states (such as sleep) are largely unregulated. Surprisingly, our data point to a third possibility: homeostatic mechanisms are implemented in neocortical circuits so as to maintain a single firing rate set point across sleep-wake states. Although we found differences in the pattern of firing across ensembles of neurons at the transitions between sleep and wake, firing rates averaged over many bouts of sleep or interspersed active wake were not significantly different. These results are consistent with one report in hippocampus (Hirase et al., 2001), while another report found small differences in average neocortical firing rates between end of wake and end of sleep (Vyazovskiy et al., 2009), and a third found larger differences in neocortical firing (∼50%) when comparing maze running

to subsequent sleep in a sleep box (Vijayan et al., 2010). Notably, the later ALOX15 two studies averaged activity over much shorter periods of time and did not control for possible circadian or environmental effects on firing. Our data show that when these factors are controlled, average V1 firing rates are conserved across sleep-wake states and suggest that a single homeostatic set point can be used to regulate activity in both states. Further, both states exhibited the same magnitude and timing of homeostatic restoration of average firing. This demonstrates that the mechanisms that restore firing in V1 can constrain the average firing of networks as they switch rapidly between very different conditions of sensory and modulatory drive.